Faculty Publications
Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736
Publications by NITK Faculty
Browse
Search Results
Item Modelling and multi-objective optimisation of squeeze casting process using regression analysis and genetic algorithm(Taylor and Francis Ltd. michael.wagreich@univie.ac.at, 2016) Gowdru Chandrashekarappa, G.C.; Krishna, P.; Parappagoudar, M.B.In the present work, an attempt has been made using statistical tools to develop a non-linear regression model and to identify the significant contribution of squeeze cast process parameters on surface roughness, hardness and tensile strength. Microstructure examination performed on the squeeze cast samples has revealed that a maximum of 100 MPa pressure is good enough to eliminate all possible casting defects. Accuracy of the developed models has been tested with the help of ten test cases. It is important to note that the developed models predict responses with a reasonably good accuracy and the developed mathematical input–output relationship helps the foundry-man to make better predictions. The present work comprises four objectives, which are conflicting in nature. Hence, mathematical formulation is used to convert four objective functions into a single objective function. The popular evolutionary algorithm, that is genetic algorithm has been utilised to determine the optimal process parameters. © 2015 Engineers Australia.Item Multi-Objective Optimization of Squeeze Casting Process using Genetic Algorithm and Particle Swarm Optimization(De Gruyter Open Ltd peter.golla@degruyter.com, 2016) Gowdru Chandrashekarappa, G.C.; Krishna, P.; Vundavilli, P.R.; Parappagoudar, M.B.The near net shaped manufacturing ability of squeeze casting process requiresto set the process variable combinations at their optimal levels to obtain both aesthetic appearance and internal soundness of the cast parts. The aesthetic and internal soundness of cast parts deal with surface roughness and tensile strength those can readily put the part in service without the requirement of costly secondary manufacturing processes (like polishing, shot blasting, plating, hear treatment etc.). It is difficult to determine the levels of the process variable (that is, pressure duration, squeeze pressure, pouring temperature and die temperature) combinations for extreme values of the responses (that is, surface roughness, yield strength and ultimate tensile strength) due to conflicting requirements. In the present manuscript, three population based search and optimization methods, namely genetic algorithm (GA), particle swarm optimization (PSO) and multi-objective particle swarm optimization based on crowding distance (MOPSO-CD) methods have been used to optimize multiple outputs simultaneously. Further, validation test has been conducted for the optimal casting conditions suggested by GA, PSO and MOPSO-CD. The results showed that PSO outperformed GA with regard to computation time. © 2016 G.C.M. Patel et al., published by De Gruyter Open.Item An intelligent system for squeeze casting process—soft computing based approach(Springer London, 2016) Gowdru Chandrashekarappa, G.C.; Krishna, P.; Parappagoudar, M.B.The present work deals with the forward and reverse modelling of squeeze casting process by utilizing the neural network-based approaches. The important quality characteristics in squeeze casting, namely surface roughness and tensile strength, are significantly influenced by its process variables like pressure duration, squeeze pressure, and pouring and die temperatures. The process variables are considered as input and output to neural network in forward and reverse mapping, respectively. Forward and reverse mappings are carried out utilizing back propagation neural network and genetic algorithm neural network. For both supervised learning networks, batch training is employed using huge training data (input-output data). The input-output data required for training is generated artificially at random by varying process variables between their respective levels. Further, the developed model prediction performances are compared for 15 random test cases. Results have shown that both models are capable to make better predictions, and the models can be used by any novice user without knowing much about the mechanics of materials and the process. However, the genetic algorithm tuned neural network (GA-NN) model prediction performance is found marginally better in forward mapping, whereas BPNN produced better results in reverse mapping. © 2016, Springer-Verlag London.